From the Southern Ocean to the North Atlantic in the Ekman Layer?

Abstract

Since the Southern Ocean encompasses the entire circumference of the globe, the zonal integral of the pressure gradient vanishes implying that the (meridional) geostrophic mass flux is zero. Conventional wisdom has it that, in view of this, the northward Ekman flux there must somehow find its way to the northern oceans, sink to the bottom (due to cooling) and return southward either below the topography or along the western boundary. Using recent (process oriented) numerical simulations and a simple analytical model, it is shown that most of the Ekman flux in the Southern Ocean does not cross the equator, nor does it sink in the northern oceans. Rather, the water that constitutes the link between the Southern Ocean and the deep water formation in the Northern Hemisphere originates in the eastern part of the southern Sverdrup interior. The associated path which takes the water from one hemisphere to the other resembles the letter “S”, where the top of the letter corresponds to the sinking region in the Northern Hemisphere and the bottom to the origin in the Southern Ocean. Although it is true that the amount of water that is crossing the equator is equal to the integrated Ekman flux in the northernmost part of the Southern Ocean, it is merely the amount (and not the origin of the water) that is equal in these two cases. The width of the transhemispheric current in the south is where t is the wind stress, ?t/?y the curl of the wind, ß the familiar variation of the Coriolis with latitude, f0 the mean Coriolis parameter, and L is the width of the basin

Southern Hemisphere westerly wind changes during the Last Glacial Maximum: model-data comparison

“…In addition to this primary ocean-CO 2 modulation mechanism, other potential oceanic impacts of wind changes have been identified. The impact of SH winds on North Atlantic Deep Water (NADW) formation through dynamical processes has been established Samuels, 1993, 1995;Rahmstorf and England, 1997;Nof and De Boer, 2004;De Boer and Nof, 2005;De Boer et al, 2008). Another hypothesis that has recently gained prominence is the suggestion that SH westerly winds affect NADW formation though its modulation of Agulhas Leakage (Sijp and England, 2009;Beal et al, 2011;Caley et al, 2012), the idea being that an equatorward shift in the SH winds forces a similar shift in the Subtropical Front, in turn reducing the amount of high salinity Indian Ocean water that enters the Atlantic Ocean.…”

The island wind-buoyancy connection

“…Consider further that, if it is strictly the Ekman layer water that represents the upper limb of the meridional overturning cell, then the sinking in the Pacific must necessarily be twice as strong as in the Atlantic, a fact which we know to be false. This issue is discussed in detail in Nof (2003) and Nof and De Boer (2004).…”

“…Some of the sinking may occur in the South Atlantic (SA) and South Pacific in the Deacon cell (Döös and Webb, 1994; Döös, 1994), but a significant portion sinks in the North Atlantic (NA), where known regions of deep convection exist. It is important to realize that it is not the southern Ekman water itself that crosses the equator and later sinks to the deep ocean in the NA, but rather the Sverdrup interior water (Nof, 2003; Nof and De Boer, 2004; see Fig. 1).…”

“… The three‐dimensional path of SO water to the northern basin. Even though the total amount of water transported northward in the SO is identical to the Ekman flux there, most of the water that constitutes the S‐shaped path corresponds to interior geostrophic flow along the eastern boundary and not to the southern Ekman flux (reproduced from Nof and De Boer, 2004). …”

The island wind–buoyancy connection

“…The three-dimensional path of SO water to the northern basin. Even though the total amount of water transported northward in the SO is identical to the Ekman flux there, most of the water that constitutes the S-shaped path corresponds to interior geostrophic flow along the eastern boundary and not to the southern Ekman flux (reproduced from Nof and De Boer, 2004).…”